DATABRICKS-SPARK Question 54

Community Discussion

Written by other learners. Comments can disagree with the marked answer — check the official documentation when in doubt.

• Chose C0

  - Wide transformations (groupBy+agg, join, distinct, repartition) require redistributing rows across partitions — a shuffle. Narrow transformations (filter, select, withColumn, map) don't. Shuffles are the most expensive part of Spark execution.